Perceptual learning of motion discrimination by mental imagery

Perceptual learning of motion discrimination by mentalimagery

Elisa M. Tartaglia $

Laboratory of Psychophysics, Brain Mind Institute,

Ecole Polytechnique Federale de Lausanne (EPFL),

Switzerland

Laboratory of Neurophysics and Physiology,

Universite Rene Descartes, Paris, France

Laura Bamert



Switzerland

Institut de Psychologie, Universite de Lausanne (UNIL),

Switzerland

Michael H. Herzog



Switzerland

Fred W. MastDepartment of Psychology, University of Bern,

Switzerland

Perceptual learning can occur when stimuli are only imagined, i.e., without proper stimulus presentation. For example,perceptual learning improved bisection discrimination when only the two outer lines of the bisection stimulus were presentedand the central line had to be imagined. Performance improved also with other static stimuli. In non-learning imageryexperiments, imagining static stimuli is different from imagining motion stimuli. We hypothesized that those differences alsoaffect imagery perceptual learning. Here, we show that imagery training also improves motion direction discrimination.Learning occurs when no stimulus at all is presented during training, whereas no learning occurs when only noise ispresented. The interference between noise and mental imagery possibly hinders learning. For static bisection stimuli, thepattern is just the opposite. Learning occurs when presented with the two outer lines of the bisection stimulus, i.e., with onlya part of the visual stimulus, while no learning occurs when no stimulus at all is presented.

Keywords: perceptual learning, mental imagery, imagery learning, motion direction discrimination

Citation: Tartaglia, E. M., Bamert, L., Herzog, M. H., & Mast, F. W. (2012). Perceptual learning of motion discrimination bymental imagery. Journal of Vision, 12(6):14, 1–10, http://www.journalofvision.org/content/12/6/14, doi:10.1167/12.6.14.

Introduction

Perceptual learning is the ability to improve percep-tion through practice (for reviews see Fahle & Poggio,2003; Fahle, 2005). It is usually assumed that percep-tual learning is stimulus-driven, i.e., the synapticmodifications underlying learning are triggered by thepresentation of the stimuli. However, we have recentlyshown that visual perceptual learning can also occur inthe absence of proper stimulus presentation whenobservers imagined the relevant part of a bisectionstimulus (Tartaglia et al., 2009b; see also Amitay,Irwin, & Moore, 2006; Dupuis-Roy & Gosselin, 2007).

A bisection stimulus consists of two outer lines, thedistance between which is bisected by a central line intwo unequal components (Supplementary MaterialFigure S1). In standard perceptual learning experi-ments, observers train to discriminate whether thiscentral line is closer to the left or right outer line.Observers’ performance in this task significantlyimproves (Crist, Kapadia, Westheimer, & Gilbert,1997; Otto, Herzog, Fahle, & Zhaoping, 2006; Tarta-glia, Aberg, & Herzog, 2009a). In a mental imageryexperiment, observers were presented only with the twoouter lines of the bisection stimulus and were asked toimagine the central line offset either to the right or tothe left. After such imagery training, observers’

Journal of Vision (2012) 12(6):14, 1–10 1http://www.journalofvision.org/content/12/6/14

doi: 10 .1167 /12 .6 .14 ISSN 1534-7362 � 2012 ARVOReceived May 16, 2011; published June 12, 2012

mailto:[email protected]

mailto:[email protected]

http://www.journalofvision.org/content/suppl/2012/06/08/12.6.14.DC1/JOV-12-6-14_Supplementary_Figures.pdf


performance, measured with the complete bisectionstimulus, improved, i.e., observers were better indiscriminating the direction of the central line (Tarta-glia, Bamert, Mast, & Herzog, 2009b).

Mental imagery is a composite of—at least partial-ly—distinct subabilities (see for example Mast &Kosslyn, 2002). Exposure to parts of static stimulioften improves mental imagery of the entire staticstimuli. For example, mentally completing a fragment-ed figure is better than imagining the entire figure (e.g.,Intons-Peterson, 1981; see also Nielsen & Smith, 1973;Tversky, 1975; Thompson & Klatzky, 1978). Thepresentation of a visual grid has also been shown tosupport mental imagery tasks (Spivey & Geng, 2001).Hence, a perceptual assistance can aid mental imageryof static stimuli (e.g., Mast, Berthoz, & Kosslyn, 2001).Unlike mental imagery of static stimuli, mental imageryof motion stimuli can be disrupted by the presentationof incongruent stimuli. For example, a motion afteref-fect induced by moving stimuli interferes with themental rotation of a letter (Corballis & McLaren,1982). Similarly, motion adaptation can interfere withimagery of a moving dot (Gilden, Blake, & Hurst,1995). Here, we show that imagery perceptual learningof static and moving stimuli shows similarly differentcharacteristics.

General methods

General setup

Stimuli were presented on a Philips 201B4 CRTmonitor driven by a RadeOn 9200 SE graphics card.The display was linearized through look-up tables andhad a luminance resolution of 10 bits. The screen wasrefreshed at 100 Hz and had a spatial resolution of 1024· 768 pixels. The room was dimly illuminated (0.5 lux)and background luminance measured on the screen wasbelow 1 cd/m2. Participants observed the stimuli from adistance of 2 m.

Observers

30 naive students from the EPFL (Ecole Polytechni-que Federale de Lausanne) and UNIL (University ofLausanne) participated in this study (age range 18–30).Students were paid for their participation. The Frei-burg visual acuity test was used to determine observers’visual acuity (Bach, 1996). Observers had to reach avalue of 1.0 (corresponding to 20/20) for at least oneeye. Observers signed informed consent and were toldthey could quit the experiment at any time they wished.Observers were naive as to the goals of the experiment.

Stimuli and task

The motion stimulus used in all experiments todetermine pre- and post-training baseline performanceconsisted of 300 dots randomly distributed within asquare field (108 width · 108 height of visual angle at aviewing distance of 2 m). Each dot subtended 5 min ofarc. Dots were white (20 cd/m2) on a dark backgroundand moved with a constant speed of 5.08/s. Dot lifetimewas 100 ms. The direction of motion of each dot waskept constant during its lifetime. When the lifetime of adot expired, a new dot appeared in the subsequentframe at a random position. The age of each dot wasrandomly chosen to avoid the dots all expiringsimultaneously. The independent variable was thepercentage of dots moving in the same direction eitherhorizontally (right/left) or vertically (upwards/down-wards). Each trial started with a red fixation dot (8 minof arc, 10 cd/m2) presented in the center of the screen,followed after 500 ms by the stimulus (300 movingdots) for 5 s. The fixation point did not disappearduring stimulus presentation. Between response andsubsequent trial, only the fixation point was presentedon the screen. A block consisted of 80 trials. In a binarytask, observers had to discriminate the direction of dotmotion (right/left or up/down). Responses were givenby pressing one of two buttons. Auditory feedback wasgiven for incorrect responses and omissions.

Procedure

The procedure was the same for all six experiments,except for the training phase.

1. Familiarization. Observers were first familiarizedwith the experimental setup by performing 80 trialsof a vertical chevron discrimination task.

2. Adaptive procedure for threshold determination. Weran 2 blocks of 80 trials (one for the horizontalmotion direction and one for the vertical motiondirection) using an adaptive procedure (PEST,Taylor & Creelman, 1967) to find the individualmotion coherence level for which observers per-formed at 75% correct (threshold).

3. Pretraining baseline measurements. We determinedbaseline performance for the horizontal and verticalmotion direction. A fixed motion coherence level wasused for each observer individually. The level was setto the threshold as determined in the previous step.This method ensures that performance is not atceiling or floor in the prebaseline measurements. Thetwo conditions were measured twice for eachobserver in a randomized order to reduce theinfluence of learning effects in the average data.

4. Extra payment. To keep observers alert during thetraining, we offered them an extra payment of 5

Journal of Vision (2012) 12(6):14, 1–10 Tartaglia, Bamert, Herzog, & Mast 2

CHF for every 2% of improvement. We informedobservers about the extra payment after havingmeasured the pretraining baselines. The extrapayment was offered to all participants in all sixexperiments.

5. Training phase. In Experiment 1, observers trainedto discriminate the direction of dot motion (right-ward vs. leftward) embedded in randomly movingdots. This was the only experiment in whichobservers trained with the proper motion stimuli.The training phase comprised a total number of4160 trials (52 blocks of 80 trials each) performed in10 separate daily sessions. On the first and last dayof training, observers performed just two blocks oftraining, i.e., 160 trials; in addition, the baselinemeasurements were conducted. Starting from thesecond day of training up to the ninth, observersperformed six blocks of training per session, i.e., 480trials per day. Experiments 2, 3, 4, and 5 (the mentalimagery experiments) involved different mentalimagery training conditions in which none or onlyparts of the motion stimulus of Experiment 1 werepresented on the screen.

Before imagery training, we presented a randomdot motion display in which the percentage ofcoherent motion to the right or to the left wassuprathreshold, yielding more than 95% correctresponses. A high frequency tone (1 kHz) wasassociated with motion direction to the right while alow frequency tone (700 Hz) was associated withmotion to the left. We told observers about theassociation which was rehearsed in 80 trials (therewas no association phase in Experiment 5). Duringimagery training, the high (1 kHz) or the low (700Hz) frequency tone indicated to imagine motion tothe right (high tone) or to the left (low tone).Observers were instructed to imagine a smallpercentage of dots moving and to push one out oftwo buttons accordingly. The training phase com-prised the same number of trials as in Experiment 1(in Experiment 1 no tones were presented). Exper-iment 6 was identical to Experiment 5, but observerswere not asked to imagine any stimulus.

6. Post-training baseline measurements. In all experi-ments, post-training baseline measurements, identi-cal to the pretraining baseline measurements, weredetermined. These measurements were conductedbetween 9 and 15 days after the pretrainingbaselines.

Data analysis

In all experiments, we compared post- and pretrain-ing d0 for the horizontal and the vertical direction ofmotion. To assess learning, we computed (two-tailed,paired) t-tests (alpha ¼ 0.05) with the null hypothesis

that there is no difference between pre- and post-training d0. In addition, in Experiment 1, regressionlines were fitted to the training data of the observers(not shown). The slopes of the regression lines werecompared with the hypothesis of no improvement ofperformance (a slope of 0.0).

General results

Experiment 1: perceptual learning of motiondirection discrimination

In a standard perceptual learning experiment,observers trained with a standard coherent motiondiscrimination task (Ball & Sekuler, 1982; Ball &Sekuler, 1987; Zohary, Celebrini, Britten, & Newsome,1994; Vaina, Sundareswaran, & Harris, 1995; Liu &Vaina, 1998; Liu, 1999; Liu & Weinshall, 2000; Seitz &Watanabe, 2003; Pilly & Seitz, 2009).

Methods

Baseline performance was measured for the hori-zontal (right/left) motion discrimination task as well asfor the vertical (upwards/downwards) motion direction(Figure 1a). Following the baseline measurements,observers trained with the horizontal motion discrim-ination stimulus for 52 blocks, i.e., 4160 total trials(Figure 1b). After training, baselines were measuredagain for both directions of motion (Figure 1c). Fournaive observers participated in the experiment.

Results

We found a significant improvement of performancefor the trained motion direction (Figure 1d; horizontalmotion direction: post-pre baseline performance ¼ 1.4;p¼ 0.039; t¼ 3.495; df¼ 3). There was no improvementfor the untrained direction of motion (Figure 1d;vertical motion direction: Post-pre baseline performance¼ 0.6; p ¼ 0.2; t ¼ 1.645; df¼ 3). In the training phase,slopes of regression lines for the horizontal motiondirection were significantly different from zero (meanslope: 0.3, p-value: 0.04; results not shown). Individualdata are shown in the Supplementary Material FiguresS2, S3.

Experiment 2: imagery training—blankscreen

We investigated whether perceptual learning ofmotion direction discrimination could occur via mentalimagery.


http://www.journalofvision.org/content/suppl/2012/06/08/12.6.14.DC1/JOV-12-6-14_Supplementary_Figures.pdf#page=2



Figure 1. Experiment 1. (a) Pretraining baseline measurements. In the horizontal motion discrimination task, observers had to discriminate

the direction of coherent motion (right or left) (a.a). The light grey arrows indicate the direction of the coherently moving dots, while the

dark grey arrows indicate random motion of the remaining dots (arrows were not presented on the real display). In the vertical motion

discrimination task, observers had to discriminate the direction of coherent motion (upwards or downwards) and push a button accordingly

(a.b). (b) Training phase. Observers trained with the horizontal motion discrimination task for 4160 trials. (c) Post-training baseline

measurements (same as in a). (d) We report results in terms of d0, a criterion free measure of sensitivity. Higher values of d0 indicate

better performance. Post-training baselines were significantly higher than pretraining baselines for the trained horizontal motion. Error

bars indicate the standard error of the mean (SEM) for four observers.


Methods

We measured baseline performance for the horizon-tal motion discrimination as well as for the verticalmotion direction (Figure 1a). During the training phase(Figure 2a), observers were presented with a blankscreen only. In each trial, a high or a low frequencytone indicated to imagine approximately 30 dots, i.e.,10% of the total number of dots, moving to the right orleft, respectively. At the end of the training session, wemeasured baseline performance again for both motiondirections (Figure 1c). Four naive observers participat-ed in the experiment.

Results

Sensitivity increased for the trained motion direction(Figure 2b; horizontal motion direction: post-prebaseline performance ¼ 0.67; p ¼ 0.04; t ¼ 3.365; df ¼3). We did not find an increase in sensitivity for thevertical direction of motion (Figure 2b; vertical motiondirection: post-pre baseline performance ¼�0.06; p ¼0.8; t¼ 0.212; df¼ 3). Individual data are shown in theSupplementary Material Figures S2, S3.

Experiment 3: imagery training—noise

Experiment 2 showed that perceptual learning ofmotion stimuli can occur via imagery when only ablank screen is presented. Previous studies have shownthat the presence of incongruent motion interferes withmental imagery. Here, we show that no improvement ofperformance occurs when noise is presented instead ofa blank screen during the imagery training.

Methods

We measured baseline performance for both thehorizontal and the vertical direction of motion (Figure1a). During the training phase, randomly moving dotswere presented. Observers were asked to imagine asubset of these dots (approximately 10%, i.e., 30 dots)to move coherently to the right or left (Figure 3a). Aftertraining, baseline measurements were repeated for bothdirections of motion (Figure 1c). Five naive observersparticipated in the experiment.

Results

Sensitivity improved neither for the trained horizon-tal motion direction, nor for the untrained vertical one(Figure 3b, horizontal motion direction: Post-pre ¼�0.071; p ¼ 0.7; t ¼ 0.498; df ¼ 4. Vertical motiondirection: post-pre ¼�0.13; p ¼ 0.6; t ¼ 0.537; df ¼ 4).Individual data are shown in the SupplementaryMaterial Figures S2, S3.

Experiment 4: imagery training—static dots

Experiment 3 showed that random motion obstruct-ed imagery perceptual learning. To investigate whetherthis interference is related to the noise, we askedobservers to imagine coherent motion while presentedwith a pattern of static dots. We expected that theprocessing of static dots interfered less with imagery.

Methods

During the imagery training phase, observers wereasked to imagine a small part of the static dots(approximately 10%, i.e., 30 dots) moving coherentlyto the right or to the left, according to the tones (Figure4a). Nine naive observers participated in the experi-ment.

Figure 2. Experiment 2. (a) Training phase. A blank screen was

presented during the imagery training. Observers were asked to

imagine a small number of dots moving coherently to the left or

right, depending on a tone (arrows and dots were not shown in the

actual display). The pre- and post- baselines conditions were

exactly the same as in Experiment 1 (not shown). (b) Baseline

performance significantly improved for the trained horizontal

motion but not for the untrained vertical motion. Error bars

indicate the standard error of the mean (SEM) for four observers.







Results

No increase in sensitivity after training with bothdirections of motion was found (Figure 4b, horizontalmotion direction: post-pre ¼ 0.25; p ¼ 0.26; t ¼ 1.206;df ¼ 8, vertical motion direction: post-pre ¼ 0.22; p ¼0.1; t¼ 1.890; df¼ 8). Individual data are shown in theSupplementary Material Figures S2, S3.

Experiment 5: imagery training—congruentmoving dots

It has been shown that a stimulus compatible withthe stimulus that has to be imagined can serve as aperceptual assistance for imagery. For example, Mastet al. (2001) have shown that a mental image of staticdots was more easily formed when some of the dots

were presented during imagery. Here, we tested

whether also imagery learning benefits from a percep-

tual assistance.

Methods

In the imagery training, observers were presented

with three dots moving to the right or left randomly

positioned on the screen. The direction of motion was

randomized across trials and was associated with a high

or a low frequency tone. Observers were instructed to

imagine a small proportion of dots (approximately 30

dots) moving in the same direction and at the same

speed as the three dots (Figure 5a). Four naive

observers participated in this experiment.

Figure 3. Experiment 3. (a) Training phase. Randomly moving

dots were presented during the imagery training. Observers were

asked to imagine a small proportion of the dots moving coherently

to the left or right, depending on a tone. (b) Baseline performance

did not improve for both the trained (horizontal) and untrained

(vertical) motion directions. Error bars indicate the standard error

of the mean (SEM) for five observers.

Figure 4. Experiment 4. (a) Training phase. A number of fixed dots

were presented during imagery training. Observers were asked to

imagine a small proportion of the dots moving coherently to the

left or to the right, depending on a tone. Pre- and post- baselines

were exactly the same as in the experiments before. (b) Baseline

performance did not improve for either the horizontal or for the

untrained, vertical motion direction. Error bars indicate the

standard error of the mean (SEM) for nine observers.




Results

Sensitivity significantly improved for the trainedmotion direction, indicating that the visual cuesefficiently guided mental imagery (Figure 5b, horizontalmotion direction: post-pre¼0.35; p¼0.045; t¼3.189; df¼3). This improvement did not transfer to the untrainedmotion direction (Figure 5b, vertical motion direction:post-pre¼ 0.4; p¼ 0.5; t¼ 0.644; df¼ 3). Individual dataare shown in the Supplementary Material Figures S2,S3.

Experiment 6: no mental imagery—movingdots

In the previous experiment, three coherently movingdots guided imagery learning. Here, as a controlexperiment, we show that motion discrimination didnot improve when observers did not imagine motion.

Methods

The training phase was identical to that of Exper-iment 5, except that observers did not have to imaginemoving dots. Observers had to discriminate the motiondirection of the three dots (Figure 6a). Four naiveobservers participated in the experiment.

Results

Performance was above 95% correct for the motiondirection task during training (results not shown). Wedid not find an increase in sensitivity for the trained aswell as the untrained motion direction (Figure 6b,horizontal motion direction: post-pre¼0.03; p¼0.7; t¼0.361; df¼3; vertical motion direction: post-pre¼0.4; p¼ 0.2; t ¼ 1.740; df ¼ 3). Individual data are shown inthe Supplementary Material Figures S2, S3.

Figure 5. Experiment 5. (a) Training phase. Three coherently

moving dots were presented during the imagery training.

Observers were asked to imagine a small proportion of dots

moving in the same direction as the three dots. (b) Baseline

performance significantly improved for the horizontal motion

direction but not for the untrained vertical direction. Error bars

indicate the standard error of the mean (SEM) for four observers.

Figure 6. Experiment 6. (a) Training phase. As in Experiment 5,

three coherently moving dots were presented during training.

Observers indicated the motion direction. No imagery was

involved in this experiment. (b) Baseline performance did not

improve through training neither for the horizontal nor for the

untrained, vertical motion direction. Error bars indicate the

standard error of the mean (SEM) for four observers.






Discussion

Most models of perceptual learning assume that thesynaptic changes underlying perceptual learning aredriven by the presentation of stimuli. Without stimuluspresentation learning is impossible. Other factors, suchas attention, motivation, or task instructions can onlymodulate learning but they cannot drive it. However,we have previously shown that perceptual learning canoccur in the absence of (proper) stimulus presentationwhen observers were asked to imagine the stimuli orparts of them (Tartaglia et al., 2009b).

Imagery of static and motion stimuli shows differentcharacteristics. While imagery of static stimuli activatesthe primary visual cortex (Kosslyn, Thompson, Kim, &Alpert, 1995; Slotnick, Thompson, & Kosslyn, 2005,Thompson, Kosslyn, Sukel, & Alpert, 2001), imageryof motion stimuli activates motion areas like MT,nearly as strongly as real motion, but not V1 (Goebel,Khorram-Sefat, Muckli, Hacker, & Singer, 1998;Grossman & Blake, 2001; Seurinck, de Lange, Achten,& Vingerhoets, 2011). This is consistent with the ideathat the activation of specific brain regions depends onthe content of the mental image (O’Craven &Kanwisher, 2000). For imagery of motion stimuli,stimuli that are incompatible with the direction of theimagined motion cause disruption of imagery (Corbal-lis & McLaren, 1982; Gilden et al., 1995). Given thesedifferences between the imagery of static and motionstimuli, we hypothesized and found that imageryperceptual learning of static stimuli is different fromimagery perceptual learning of motion stimuli.

We found that motion discrimination improvedwhen observers imagined left-right moving dots on ablank background during training (Figure 2). To thecontrary, in the case of mental imagery of bisectionstimuli no improvement of performance occurred whenonly a blank screen was provided (see SupplementaryMaterial Figure S1 and Tartaglia et al., 2009b).Possibly this discrepancy comes from the fact that highspatial resolution mental images, required for bisectionstimuli, need a spatial reference. On the other hand,imagery perceptual learning of Gabor patch detectionis possible with no stimulus presentation duringtraining because no spatial reference is needed (Tarta-glia et al., 2009b). Such a reference seems also not to benecessary for motion imagery learning. However,motion discrimination failed to improve when theimagined motion was incompatible with motion of realstimuli (Figure 3). Such disruptive interference wasobserved also when static dots were presented duringmotion imagery (Figure 4). Interestingly, Experiment 3was the only experiment in which observers reporteddifficulties in performing the mental imagery task,indicating that, indeed, the randomly moving dots wereimpeding imagery. Accordingly, Experiment 5 shows

that when stimuli were congruent with the mentalimage, they could serve as a guide or a perceptualassistance to mental imagery (Mast et al., 2001),yielding improvement of performance (Figure 5).

As mentioned above, imagery bisection learningworks only if the outer lines are presented. It may beargued that similarly to motion imagery, static imageryneeds perceptual assistance. However, while motionimagery benefits from the perceptual assistance of asmall signal which varies from trial to trial, staticimagery rather needs a spatial reference which remainsconstant from trial to trial and does not contain anyinformative signal.

All together, the current results provide yet anotherdemonstration that mental imagery can stand in for thevisual stimulus, and that imagery training improvesvisual perception.

Data for the trainedmotion direction are rather clearand consistent for conditions where performanceimproved (Experiments 1, 2, and 5) and whereperformance remained virtually unchanged (Experi-ments 3 and 6). An ANOVA with a single factor (factorExperiments) showed that Experiment 3 was signifi-cantly different from imagery learning with a blankscreen (p ¼ 0.039). Results for the untrained motiondirection are less clear. Data for these conditions are ingeneral noisy and the associated p-values vary greatly.The scatter plots reported in the SupplementaryMaterial (Figure S2) clearly show that high p-valuescome from high interobserver variances. We havedetermined performance for the orthogonal motiondirection because it is standard in perceptual learningexperiments to investigate transfer of learning tountrained stimulus features, and in a previous manu-script on imagery learning of static stimuli (Tartaglia etal., 2009b), contrary to most findings in standardperceptual learning, there was transfer to the untrainedorientation of bisection and Gabor stimuli. The reasonsfor this transfer remain unclear for the moment. Itmight be that the small number of trials per trainingsession causes transfer, as shown in Aberg, Tartaglia, &Herzog (2009). Also the present data are not conclu-sive. It might be argued that the improvement ofperformance in the untrained motion direction is anindication of rather unspecific learning effects whichare not related to imagery. In our previous manuscript,we showed, however, that learning is indeed caused byimagery. For example, no learning occurred whenobservers were presented with the two outer lines of thebisection stimuli, as during imagery training, but wererequired to perform a different task which did notinvolve imagery (Tartaglia et al., 2009b). In thiscontribution, we also showed that learning occurredwhen observers imagined motion on top of a very smallsignal (Experiment 5). No learning occurred when





observers were presented with the very same smallsignal but were not required to imagine (Experiment 6).

The improvement of performance and the lackthereof do not have a straightforward interpretationin terms of initial threshold or initial pre-trainingperformance, i.e., we found no consistent pattern ofcorrelation between learning and initial threshold orinitial performance (see Supplementary Material Fig-ures S4, S5).

In summary, imagery perceptual learning of motionstimuli is possible, does not require any visual input,and is disrupted by the presence of incompatible noise.Imagery of static and motion stimuli shows differentcharacteristics and so does imagery learning of staticand motion stimuli. Accordingly, it is very likely thatimagery learning relies on the same mechanisms activeduring imagery.

Acknowledgments

We would like to thank Bilge Sayim for helpfuldiscussions and Marc Repnow for technical support.Elisa Tartaglia is funded by the Sinergia project ‘‘Staterepresentation in reward based learning in humanhealthy observers, schizophrenic patients, and modelsof perceptual learning’’ and by the "Perspective Re-searcher fellowship" (project number PBELP3-135838)of the Swiss National Science Foundation (SNF).

Commercial relationships: none.Corresponding author: Elisa Tartaglia.Email: [email protected]: Laboratory of Neurophysics and Physiology,UMR 8119 CNRS, Universite Rene Descartes, Paris,France.

References

Aberg, K. C., Tartaglia, E. M., & Herzog, M. H.(2009). Perceptual learning with chevrons requires aminimal number of trials, transfers to untraineddirections, but does not require sleep. VisionResearch, 49, 2087–2094.

Amitay, S., Irwin, A., & Moore, D. R. (2006).Discrimination learning induced training withidentical stimuli. Nature Neuroscience, 9(11),1446–1448.

Bach, M. (1996). The Freiburg visual acuity test—Automatic measurement of visual acuity. Optome-try and Vision Science, 73, 49–53.

Ball, K., & Sekuler, R. (1982). A specific and enduring

improvement in visual motion discrimination.Science, 218, 697–698.

Ball, K., & Sekuler, R. (1987). Direction-specificimprovement in motion discrimination. VisionResearch, 27, 953–965.

Corballis, M. C., & McLaren, R. (1982). Interactionbetween perceived and imagined rotation. Journalof Experimental Psychology: Human Perception andPerformance, 8(2), 215–224.

Crist, R. E., Kapadia, M. K., Westheimer, G., &Gilbert, C. D. (1997). Perceptual learning of spatiallocalization: Specificity for orientation, position,and context. Journal of Neurophysiology, 78(6),2889–2894.

Dupuis-Roy, N. & Gosselin, F. (2007). Perceptuallearning without signal. Vision Research, 47(3),349–356.

Fahle, M., & Poggio, T. (2003). Perceptual learning.Boston: MIT Press.

Fahle, M. (2005). Learning to tell apples from oranges.Trends in Cognitive Sciences, 9, 455–457.

Gilden, D., Blake, R., & Hurst, G. (1995). Neuraladaptation of imaginary visual motion. CognitivePsychology, 28(1), 1–16.

Goebel, R., Khorram-Sefat, D., Muckli, L., Hacker,H., & Singer, W. (1998). The constructive nature ofvision: Direct evidence from functional magneticresonance imaging studies of apparent motion andmotion imagery. European Journal of Neuroscience,10, 1563–1573.

Grossman, E. D., & Blake, R. (2001). Brain activityevoked by inverted and imagined biological mo-tion. Vision Research, 41(10-11), 1475–1482.

Intons-Peterson, M. J. (1981). Constructing and usingunusual and common images. Journal of Experi-mental Psychology: Human Learning and Memory,7, 133–144.

Kosslyn, S. M., Thompson, W. L., Kim, I. J., & Alpert,N. M. (1995). Topographical representations ofmental images in primary visual cortex. Nature,378, 496–498.

Liu, Z. (1999). Perceptual learning in motion discrim-ination that generalizes across motion directions.Proceedings of the National Academy of Sciences,96(24), 14085–14087.

Liu, Z., & Vaina, L. M. (1998). Simultaneous learningof motion discrimination in two directions. Cogni-tive Brain Research, 6(4), 347–349.

Liu, Z., & Weinshall, D. (2000). Mechanisms ofgeneralization in perceptual learning. Vision Re-search, 40, 97–109.





Mast, F. W., Berthoz, A., & Kosslyn, S. M. (2001).Mental imagery of visual motion modifies theperception of roll-vection stimulation. Perception,30, 945–957.

Mast, F. W., & Kosslyn, S. M. (2002). Visual mentalimages can be ambiguous: Insights from individualdifferences in spatial transformation abilities. Cog-nition, 86(1), 57–70.

Nielsen, G. D., & Smith, E. E. (1973). Imaginal andverbal representations in short-term recognition ofvisual forms. Journal of Experimental Psychology,101(2), 375–378.

O’Craven, K. M., & Kanwisher, N. (2000). Mentalimagery of faces and places activates correspondingstimulus-specific brain regions. Journal of CognitiveNeuroscience, 12(6), 1013–1023.

Otto, T. U., Herzog, M. H., Fahle, M., & Zhaoping, L.(2006). Perceptual learning with spatial uncertain-ties. Vision Research, 46, 3223–3233.

Pilly, P. K., & Seitz, A. R. (2009). What a difference aparameter makes: A psychophysical comparison ofrandom dot motion algorithms. Vision Research,49(13), 1599–1612.

Seitz, A. R., & Watanabe, T. (2003). Psychophysics: Issubliminal learning really passive? Nature, 422–436.

Seurinck, R., de Lange, F. P., Achten, E., &Vingerhoets, G. (2011). Mental rotation meets themotion aftereffect: The role of hV5/MTþ in visualmental imagery. Journal of Cognitive Neuroscience,23(6), 1395–1404.

Slotnick, S. D., Thompson, W. L., & Kosslyn, S. M.(2005). Visual mental imagery induces retinotopi-cally organized activation of early visual areas.Cerebral Cortex, 15, 1570–1583.

Spivey, M. J., & Geng, J. J. (2001). Oculomotormechanisms activated by imagery and memory: Eyemovements to absent objects. Psychiatry Research,65(4), 235–241.

Tartaglia, E. M., Aberg, K. C., & Herzog, M. H.(2009a). Perceptual learning and roving: Stimulustypes and overlapping neural populations. VisionResearch, 49, 1420–1427.

Tartaglia, E. M., Bamert, L., Mast, F. W., & Herzog,M. H. (2009b). Human perceptual learning bymental imagery. Current Biology, 19, 2081–2085.

Taylor, M., & Creelman, C. (1967). Pest: Efficientestimates on probability functions. The Journal ofthe Acoustical Society of America, 41, 782–787.

Thompson, A. L., & Klatzky, R. L. (1978). Studies ofvisual synthesis: Integration of fragments intoforms. Journal of Experimental Psychology: HumanPerception and Performance, 4(2), 244–263.

Thompson, W. L., Kosslyn, S. M., Sukel, K. E., &Alpert, N. M. (2001). Mental imagery of high- andlow-resolution gratings activates area 17. Neuro-image, 14(2), 454–464.

Tversky, B. (1975). Pictorial encoding of sentences insentence-picture comparison. Quarterly Journal ofExperimental Psychology, 27, 405–410.

Vaina, L. M., Sundareswaran, V., & Harris, J. G.(1995). Learning to ignore: Psychophysics andcomputational modeling of fast learning of direc-tion in noisy motion stimuli. Cognitive BrainResearch, 2(3), 155–163.

Zohary, E., Celebrini, S., Britten, K. H., & Newsome,W. T. (1994). Neuronal plasticity that underliesimprovement in perceptual performance. Science,263, 1289–1292.


Perceptual learning of motion discrimination by mental imagery

Documents

Transcript of Perceptual learning of motion discrimination by mental imagery